Cerebral Abscess

Cerebral Abscess: What’s the Data?

Lee Selznick, M.D. March 23rd, 2005

Cerebral Abscess: The Data 800

A rtic le s

700 600 500 400 300 200 100 0 19021965

19661970

19711975

19761980

Total

19811985

19861990

English

19911995

19962000

CCJ

Total Articles (1902-2005): 4541 In English Language: 2957 In CCJ: 697

20012005

*Medline

Cerebral Abscess • Epidemiology – Risk factors – Bacteriology

• Diagnostics (Radiology/Pathology) – Diagnosing a ring-enhancing lesion – Staging an abscess

• Treatment – Surgical versus non-surgical – Steroids

Epidemiology 2500

2229

Articles

2000

1443

1500 1000

351

500 0 Total

English

CCJ

Medline Search: • Subject: cerebral abscess • Limit to: epidemiology, etiology, microbiology • Data: case series, case reports, reviews

Epidemiology Largest series: “Brain abscess: a review of 400 cases” - J Neurosurg 55:794-799, 1981 (China) China

Largest series in U.S.: “Trends in the management of bacterial brain abscesses: A review of 102 cases over 17 years” - Neurosurgery 23(4):451-458, 1988 (Rosenblum, UCSF) UCSF

Most current large series: “Bacterial brain abscess: microbiological features, epidemiological trends and therapeutic outcomes” - QJM 95: 501-509, 2002 (Taiwan, 123 cases) cases

Epidemiology •Incidence: 1500-2500 cases/year in the U.S. •1-2% of intracranial space-occupying lesions in the U.S. •Age/Sex: M > F (1.5-3:1) •Risk factors: (Mampalam & Rosenblum, UCSF, 1970-1986) Out of 96 patients with cerebral abscess: • 21% without known risk factor/source • 19% with local infection (sinusitis/mastoiditis) • 18% with cardiac source • congenital cyanotic heart disease (children) • 17% with prior intracranial surgery • 9% with prior cranial trauma • 7% with pulmonary source • A-V shunts (adults, ROW syndrome) • 5% with immunosuppression (HIV, transplant)

•Source:

Epidemiology

• Local (@40%) • middle ear/mastoiditis to temporal lobe or cerebellum • frontal sinus to frontal lobe • trauma/surgery • single > multiple • Hematogenous (@40%) • multiple • distal MCA, gray-white junction • Adult: pulmonary fistula (HHT) • Peds: cyanotic heart disease • Unknown (@20%)

Epidemiology

• Microbiology (Mampalam & Rosenblum, UCSF, 1970-1986) • 55% Single organism • 75% Aerobic (streptococcus > staph aureus, haemophilus) • 20% Anaerobes • 20% Multiple • 25% Sterile cultures • Immunosuppressed (Neurology 50(1): 1-17, 1997 – HIV practice guidelines) • toxoplasmosis (*most common CNS mass lesion) • fungal • mycobacterium • Neonates: • proteus (J Neurosurg 69: 877-882, 1988 – review of 30 cases) • citrobacter • bacteroides

Diagnostics (Radiology/Pathology) 2500

2169

Articles

2000

1456

1500 1000

328

500 0 Total

English

CCJ

Medline Search: • Subject: cerebral abscess • Limit to: CSF, diagnosis, radiography, radionucleotide imaging, u/s • Data: case series, case reports, small case-control studies, animal studies (staging)

Ring-enhancing Lesions DDx (Osborne) • Common: •GBM •Metastasis •Abscess •Granuloma (fungal, TB) •Resolving hematoma •Subacute Infarct

• Less common: •Thrombosed AVM •Active MS plaque

•Uncommon: •Thrombosed aneurysm •Lymphoma •Radiation necrosis

Radiology – Abscess vs. Neoplasm • CT • thin, smooth ring versus nodular ring on contrast scan Abscess

GBM

ABSCESS *may be thinner on medial surface (50%) *thinner ring if immunosuppressed

• MRI (Acta Radiol suppl. 369: 754, 1986) • HYPOintense ring versus heterogeneous signal on T2

Radiology – Abscess vs Neoplasm • MR spectroscopy (AJNR 23: 1369-1377, 2002) • acetate and amino acid peaks • present in 4 of 5 abscesses • absent in 7 of 7 neoplasms

• MR diffusion (J Neurosurg 97: 1101-1107, 2002) • 16 of 16 abscesses with restricted diffusion • 16 of 16 neoplasms with unrestricted diffusion (AJNR 22: 1738-1742, 2001) •1 of 3 abscesses with unrestricted diffusion •1 of 13 neoplasms with restricted diffusion

• Leukocyte scan (radionucleotide): (Neurosurgery 16: 23-26, 1985) • 16 patients w/ mass lesion c/w tumor vs abscess (ring-enhancing) • positive in 4 of 5 abscesses (effect of antibiotics?) • negative in 10 of 11 neoplasms (extensive necrosis/inflammation?) • takes 6-24hrs to obtain imaging

Staging Britt & Enzmann (1979-1983) • Dog model of cerebral abscess (J Neurosurg 55: 590-603, 1981) • 19 “cases” and 2 controls • direct injection of alpha strep into left parietal lobe • radiological and pathological correlation • Human pre-operative staging (J Neurosurg 59: 972-989, 1983) • small, non-randomized study • 14 patients, no controls • prospective (?)

Staging Stage I Early Cerebritis (day 1-3)

Pre-Contrast CT Irregular area of low density

Post-Contrast CT +/- enhancement, May be patchy or ring-like

Path perivascular polymorphonuclear infiltrate marked cerebral edema around lesion

Staging Stage II Late Cerebritis (days 4-9)

Pre-Contrast CT Large area of low density

Post-Contrast CT Typical ring enhancement May be solid if small

Path Mixed poly/mx infiltrate Maximal size of necrotic center Fibroblasts around necrotic center Maximal cerebral edema

Staging Stage III

Pre-Contrast CT

Post-Contrast CT

Early Faint ring separates low Ring enhancement Capsule density necrotic center (day 10-13) from low density May be thinner on surrounding edema ventricular side

Path Maturing collagen capsule (less developed on ventricular side) Less inflammation Vascularity around capsule at max

Staging Stage IV

Pre-Contrast CT

Late Capsule (day 14+)

Faint ring separates low density necrotic center from low density surrounding edema

Post-Contrast CT Ring enhancement (usually thin/dense) May be thinner on ventricular side

Path Completed collagen capsule Reactive gliosis with less edema

Articles

Treatment 2000 1800 1600 1400 1200 1000 800 600 400 200 0

1743 1170

249 Total

English

Medline Search: • Subject: cerebral abscess • Limit to: drug therapy, surgery, therapy • Data: case series and case reports

CCJ

History • 1876: Sir William Macewen • Pyogenic Infectious Diseases of the Brain and Spinal Cord, 1893 • First to propose operative management for brain abscess • Advocated abscess drainage • 1926: Walter Dandy (JAMA 87: 1477-1478, 1926) • First to advocate aspiration as primary treatment • 1936: C. Vincent (Gaz. Med. Fr., 43: 93-96, 1936) • First to advocate complete excision as primary treatment • 1971: Heinemann and Baude (JAMA 218: 1542-1547, 1971) • First to suggest medical management alone (cerebritis) • 1975: Chow (West. J. Med., 122: 167-171, 1975) •First non-surgical cure of encapsulated abscess (Listeria) • 1976-1980: publications advocate CT-guided stereotactic aspiration

Aspiration

• CT, MRI, or U/S guided • Goals • confirm diagnosis • identification of organisms/sensitivity • remove mass effect • remove bacterial “load” and improve local environment for Abx • Relative Indications • single or multiple abscesses • superficial or deep • cerebritis or capsule stage • after medical therapy alone if: (J Neurosurg 52: 217-225, 1980) • no decrease in size within 4wks • increase in size • decline in neurological status • Start antibiotics AFTER aspiration (Neurosurgery 23(4): 451-458, 1988) • review of 102 cases (1970-1986) • 30% versus 4% sterile cultures if abx given pre-op

Surgical Excision • Goals • confirm diagnosis • identification of organisms/sensitivity • remove mass effect • Relative Indications • single abscess • superficial, non-eloquent location • well-formed (capsule stage) • especially for: • traumatic, retained foreign body (J Neurosurg 28: 166-168, 1968) •

• recurrent infection 36 years after trauma fungal (J Neurol. Neurosurg. Psychiatr., 36: 758-768, 1973) • risk of recurrence (within capsule wall)

• multiloculated • cerebellar

Medical Therapy Indications • Black (J Neurosurg 38: 705-709, 1973) • able to culture bacteria from abscess (ave. 5cm) despite good concentration of antibiotics within cavity • all failed medical management (6 patients)

• Rosenblum (J Neurosurg 52: 217-225, 1980) • Series of 12 patients • abscesses that resolved with antibiotics alone (8) were 1.7cm vs. 4cm average for those ultimately requiring surgery (4) • no abscess larger than 2.5cm resolved without surgical intervention • first decrease in abscess size within a mean of 2.4wk (range: 1-4wks) • resolution of contrast enhancement by 3.5 months

• Rosenblum (Clin Neurosurg 33: 603-632, 1986) • literature review of nonoperative management (1975-1985) • 23 articles, 67 cases treated w/ medical therapy alone • Outcome of five largest series (50 cases total): • 74% overall success rate • 4% mortality

Medical Therapy Indications Relative Indications: • known source/organism • cerebritis > capsule stage • good neurological status • multiple abscesses • small size (1.5-3.0cm) • bleeding diathesis • severe medical comorbidities

Recommended Protocol • weekly CT first 4 weeks • then monthly until: • lack of contrast enhancement • off antibiotics for 2 weeks • rescan w/ any clinical deterioration • surgery if: • increase in size at any time • no change after 4 weeks of antibiotics • continue abx for at least 6-8weeks Obana & Rosenblum (Neurosurg Clin 3(2): 359, 1992)

Steroids: Good or Bad Background (Pathophysiology) • Chemotaxis of polys within 1 hour • continue to enter for 24hrs • remain for 1-2 days • bacteriocidal activities • Chemotaxis of monocytes/mx w/in 24hrs • continue to enter for first week • major bacteriocidal action w/in abscess • stimulate formation of capsule and angiogenesis • Steroids • impair chemotaxis of polys & monocytes/mx • impair formation of capsule • may reduce antibiotic penetration (Arch Intern Med 93: 850-861, 1954)

Steroids: Animal Models • Bohl et al. (Adv. Neurosurg 9: 235, 1981) – S. aureus abscess in cats – Steroids reduce edema, inflammation, and encapsulation

• Neuwelt et al. (J Neurosurg 61: 430-439, 1984) – E. coli abscess in rats – Steroids decrease macrophages and inhibit gliosis

• Quartey et al. (J Neurosurg 45: 301-310, 1976) – Strep and Staph abscesses in rabbits – Steroids decrease wbc access, organism killing, and encapsulation

Steroids: Human Data • Rosenblum et al. (J Neurosurg 49: 658-668, 1978) – NO relationship between presence or duration of steroids and mortality (36 patients, retrospective, non-randomized review)

• Mampalam & Rosenblum (Neurosurgery 23(4): 451-458, 1988) – Review of 102 cases (1970-1986) – Steroids correlated with poorer neurological outcome (worse initial neurological grade also)

• Rosenblum (Neurosurgery 36(1): 76-86, 1995) – Review of 16 cases with multiple abscess – Steroids did not correlate with outcome

• Takehsita et al. (Japan) Acta Neurochir 140: 1263-1270, 1998) – Review of 113 cases (1976-1995) – 24 treated with steroids (all with impaired consciousness and edema) – Steroids did not correlate with outcome

Steroids: Recommendations • Useful for reduction of symptomatic mass effect caused by edema • Should be avoided in early stages if possible • Use of steroids may decrease enhancement – If cerebritis (Britt & Enzmann, J Neurosurg 59: 972-989, 1983)

• Withdrawal of steroids may increase enhancement – Does not correlate clinically (Radiology 135: 663-671, 1980)

Morbidity/Mortality 800

727

700

Articles

600

473

500 400 300

119

200 100 0 Total

English

Medline Search: • Subject: cerebral abscess • Limit to: complications, mortality • Data: case series and case reports

CCJ

• Reduced mortality since use of CT scan • 30-50% mortality pre-CT era • 5-10% mortality post-CT era • Poor outcome highly associated with: • initial neurologic grade (J Neurosurg 55: 794-799, 1981) (review of 400 cases from China) • intraventricular rupture (Acta Neurochir : 1263-1270, 1998) • review of 113 patients from Japan • odds ratio of 24.5 (95% CI 3.04-197.9) • 39% mortality versus 3.4% • sepsis (QJM 95: 501-509, 2002) (review of 123 cases from Taiwan)

• High risk of seizures • 25% pre-operative risk (Neurosurgery 23(4): 451-458, 1988) • 30-50% in most series (Osenbach, Neurosurg Clin 3(2): 403-420, 1992) • mean latency for onset up to 3.5 years (Brain 96: 259-268, 1973) • usually well-controlled with anti-convulsants

THE END